Measurements of Data Analytics Capability Construct

Abstract

Big data analytics has emerged as an important area of research in the big data domain. However, big data analytics capability empirical investigations have been hindered by the lack of psychometrically sound measurement items and scales. This paper reports approaches to the development and validation of new multi-item measurement scales reflecting a construct called data analytics capability. Data analytics capability reflects an organization's expertise in developing and deploying resources, usually in combination, to achieve the desired new insights that have business implications. This data analytics capability competence is operationalized as a multidimensional construct reflected by measurement items. In the first stage of measure development, we review the prominent information systems journals. We develop the construct items most of which are new items. Then we use industry experts and academicians to score them on a one to five scale. We also asked them to propose any new items. Based on their evaluation and scoring we finalize four items for this construct. Our results demonstrate that a reduced set of measurement items have reasonable psychometric properties and, therefore, are useful inputs for multi-item measurement scale development. In the second stage of measurement development, we conduct a survey of big data users via two online user groups. We received 349 valid responses which are analyzed using SPSS and AMOS statistical software. We successfully performed construct reliability analysis. The construct developed in this research may be used to advance scholarly understanding and theory in the big data and data analytics field.